The consumed charge Q referred to a period Δt can be determined to perform a precalculation of the operating time expected in mobile devices, which are generally operated with a battery or an accumulator.
It will be appreciated that there are differences in power consumption during the operation of mobile stations of digital mobile radio systems. This is at least partially attributable to the existence or non-existence of radio transmission links.
There are two possibilities for calculating power consumption.
First, a current measurement and integration of power variation with time can be used to calculate power consumption. Since, as mentioned above, the power consumption is subject to great fluctuations, the current measurement can be quite complex, requiring a high-resolution analog/digital converter.
A second way to calculate power consumption, which can be applied in packet-oriented mobile radio systems such as, for example, Bluetooth, relates to storing mean power consumption for various packet types. Bluetooth has a large number of packet types for different connections (synchronous/asynchronous) and tasks (connection set-up, pollen etc.). By monitoring the packet types emitted and received, the charge consumed in the observation period can be inferred by adding together the packet-related typical power consumptions. However, this method has the several disadvantages that may affect its accuracy. First, calculating an estimated value for the actual power consumption can be very complex because the packet-typical charge consumption is composed of a number of components. For example, constant components, such as a component which is associated with a packet-linked carrier frequency change, and variable components, which depend on the amount of data transmitted per packet, must be taken into consideration. Additionally, incomplete data packets and/or a nonuniform distribution of the data over various time slots may exist when converting the power consumption from a previous data rate to a current data rate. Also, the system may not be periodic despite a constant data rate. The periodic behavior of the system can, for example, be disturbed by complex software processes or also by externally initiated events.